A larger asset can become more efficient while the system required to feed it becomes less efficient.
Executives are trained to look for economies of scale. Larger facilities can spread fixed costs, justify specialised technology, consolidate expertise and increase utilisation. Those advantages are real.
They are also incomplete.
A facility does not operate in isolation. It consumes materials, energy, labour, transport, infrastructure and supplier capacity. As the asset grows, the radius and complexity of the network required to support it may grow as well. At some point, efficiency inside the plant can be offset by diseconomy outside the plant.
The best scale for the asset is therefore not necessarily the best scale for the enterprise system.
The Strategic Context
Manandhar and Shah explored this tension in a 2017 life-cycle assessment of corn-stover feedstock logistics for a hypothetical 114 million litre-per-year cellulosic biorefinery in the US Midwest. They compared three supply configurations involving direct delivery of bales, intermediate depots, and pelletisation before final transport.
Pelletisation increased bulk density and could reduce transport burden, but it also required electricity and processing. In the scenarios studied, pellet production was a major source of additional environmental impact. The comparison changed under different electricity mixes and became more favourable to enhanced densification as supply distances increased.
The paper also highlighted a structural tension recognised in the bioenergy literature: conversion facilities can experience economies of scale while low-density biomass logistics experience diseconomies of scale because larger facilities draw material from wider areas.
The exact thresholds belong to the study's assumptions. The broader strategic insight is durable:
Scale changes the architecture of the supporting network, and the network can reverse the apparent advantage of the asset.
What Leaders Commonly Misread
The first misread is calculating scale economics inside the fence line.
A larger plant may reduce conversion cost per unit while increasing inbound kilometres, supplier concentration, inventory exposure, road congestion, warehousing requirements or vulnerability to disruption.
The second misread is treating logistics as a variable cost that scales smoothly.
Real networks contain thresholds. A larger facility may require a new depot, rail siding, port capacity, transmission connection, labour catchment or storage regime. Costs can arrive in steps rather than in a neat linear relationship.
The third misread is assuming densification or consolidation is always beneficial.
The corn-stover study shows why. Densification improves transport efficiency but adds processing energy. Whether the trade is favourable depends on distance, energy source and system configuration.
The fourth misread is using current network conditions to justify a long-lived asset.
A facility built for 25 years may face different energy prices, electricity mixes, suppliers, transport constraints and demand patterns. Scale is partly a bet on the persistence of the surrounding network.
Reframing the Issue
Facility sizing should be treated as a system architecture decision.
The relevant question is not:
"What plant size produces the lowest unit conversion cost?"
It is:
"What combination of asset scale, network design, logistics, resilience and future optionality produces the greatest enterprise value?"
That requires leadership to view the plant and its supporting ecosystem as one design problem.
Related article: Local Manufacturing or Offshore Expansion? Evaluate the System, Not the Unit Cost
Strategic Analysis: The Scale Boundary Moves
Economies inside, diseconomies outside
As assets become larger, internal capital efficiency can improve. Yet supporting networks can become longer, more concentrated and more fragile.
A large hospital can centralise expensive equipment but increase patient travel and ambulance dependency. A large data centre can concentrate technical capability while increasing grid connection risk and cooling demand. A large distribution centre can reduce warehousing overhead while increasing last-mile distance. A large defence maintenance hub can deepen specialist capability while reducing geographic redundancy.
These are not arguments against centralisation. They show that centralisation has a system cost that must be modelled.
The feedstock or demand radius is a strategic variable
For a material-processing asset, larger capacity generally requires either more supply from the same radius or a wider sourcing radius.
The first can increase competition for local feedstock. The second increases transport and dependence on a larger network.
A business case that treats supply as a fixed price per tonne can miss these dynamics. The marginal tonne may be materially more expensive and risky than the average tonne.
Supporting technologies can change the preferred scale
Densification in the Manandhar and Shah study is a good example. By increasing bulk density, pelletisation changes transport economics. But because it consumes energy, its value depends on electricity characteristics and distance.
The general principle is that enabling technologies can move the scale optimum.
Automation can reduce labour constraints. Better forecasting can reduce inventory. Renewable energy can change processing emissions. Modular equipment can reduce the penalty of decentralisation. Digital platforms can coordinate distributed operations.
Scale should therefore be reviewed when enabling technology changes materially, not treated as a once-only design assumption.
Resilience changes the economics of scale
The lowest-cost steady-state configuration may be highly concentrated.
But concentration can increase exposure to single points of failure. A disruption at one mega-facility can create a larger enterprise consequence than a disruption at one node in a distributed network.
Resilience has an economic value even when it does not improve normal operating cost. This is why portfolio leaders must resist optimisation models that exclude disruption, recovery time and strategic redundancy.
Growth can change the optimum after investment
An architecture that is sensible at one throughput can become inefficient at another.
This creates a sequencing problem. Leadership may need to choose between building large early, expanding modularly, creating distributed hubs or investing first in network capability.
The most valuable option may be the one that preserves flexibility until demand and network behaviour become clearer.
Decision Framework
ERANORTH recommends a Scale-System Fit review for major capacity decisions.
1. Define the internal scale benefit
Identify precisely what improves as the asset becomes larger: fixed-cost absorption, conversion efficiency, specialist utilisation, purchasing power, technology capability or another factor.
2. Define the network penalty
Model how larger scale changes sourcing radius, transport, inventory, infrastructure, labour and supplier concentration.
3. Identify threshold investments
Which new depots, connections, warehouses, roads, pipelines, utilities or compliance systems are triggered only beyond certain scales?
4. Model the marginal unit
Do not use only average cost. Estimate the cost, distance, carbon, risk and quality of the next increment of supply or demand.
5. Test enabling technologies
Could densification, modularity, automation, storage, digital coordination or alternative transport materially change the preferred architecture?
6. Price resilience and optionality
Assess disruption exposure, recovery time, dependence on single nodes and the value of being able to expand, contract or relocate capacity.
7. Compare architectures, not only sizes
Evaluate large centralised, distributed, hub-and-spoke, modular and phased alternatives where they are credible.
From Strategy to Execution
Immediate action: for significant asset proposals, add a supply-network sensitivity to the business case. Require the investment team to show how unit economics change when sourcing radius, transport distance and infrastructure requirements expand.
Medium-term capability building: integrate network modelling with capital planning. Operations, supply chain, finance and engineering should evaluate facility scale from the same model rather than optimise their own components independently.
Long-term strategic positioning: preserve optionality where uncertainty is high. Modular capacity, multi-source supply, distributed storage and adaptable logistics can have strategic value even if their initial unit cost is not the minimum.
Related article: The Counterfactual Is Part of the Investment Case
Signals to Monitor
Watch increasing average and marginal transport distance; supplier concentration; rising inbound variability; logistics emissions growing faster than production; depots operating near capacity; infrastructure constraints becoming critical; feedstock quality declining with wider sourcing; emergency sourcing becoming routine; and proposed capacity increases that assume unchanged input prices despite a larger supply radius.
Another important signal is divergence between plant efficiency and system efficiency. If conversion cost falls while total delivered cost, inventory or logistics risk rises, the scale decision needs to be revisited.
Questions for the Leadership Team
- Where do the economics of scale inside the asset begin to create diseconomies in the supporting network?
- What does the marginal tonne, customer, kilometre or transaction cost compared with the average?
- Which infrastructure thresholds are hidden inside the next stage of growth?
- How sensitive is the preferred architecture to electricity mix, transport cost, supplier availability or demand density?
- What resilience are we sacrificing for scale efficiency?
- Could modular or distributed capacity preserve more strategic optionality?
- At what point would we redesign the network rather than simply make the central asset larger?
Closing Perspective
Scale is not a property of the plant alone. It is a property of the whole system that makes the plant useful.
A larger facility may be economically superior in conversion terms and strategically inferior once transport, infrastructure, concentration and resilience are included. The reverse can also be true: enabling technologies can make large-scale systems increasingly attractive.
The discipline is therefore to stop asking for the "optimal plant size" in isolation.
The decision is the architecture.
Source basis: This article is an original ERANORTH synthesis principally informed by Manandhar and Shah (2017), Life cycle assessment of feedstock supply systems for cellulosic biorefineries using corn stover transported in conventional bale and densified pellet formats, Journal of Cleaner Production, volume 166. Study-specific environmental values and historical US biofuel-policy context are not presented as current benchmarks.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.